The present invention relates to a purification method for a crude PNPNH-compound.
Compounds having the general structure PNPNH are well known ligand systems which can be successfully used in a catalyst for the oligomerization of ethylene. Here, they function as ligands to be reacted with, preferably, chromium catalysts. Together with a suitable cocatalyst such a system is effective in the di-, tri- and/or tetramerization of ethylene.
For example, EP 2 239 056 B1 describes a catalyst composition and a process for the di-, tri- and/or tetramerization of ethylene. The catalyst composition comprises a chromium compound, a ligand of the general structure R1R2P—N(R3)—P(R4)—N(R5)—H and a co-catalyst acting as activator. The ligand's substituents R1, R2, R3, R4, and R5 are independently selected from a number of functional groups, comprising (among others) C1-C10-alkyl, aryl and substituted aryl. The chromium source is CrCl3(THF)3, Cr(III)acetyl acetonate, Cr(III)octanoate, Cr-hexacarbonyl, Cr(III)-2-ethylhexanoate or (benzene)tricarbonyl-chromium (wherein THF is tetrahydrofuran). The co-catalyst or activator is trimethylaluminum, triethylaluminum, triisopropylaluminum, triisobutylaluminum, ethylaluminumsesquichloride, diethylaluminum chloride, ethylaluminum dichloride, methylaluminoxane, or a combination comprising at least one of the foregoing.
A preferred choice of catalyst constituents comprises CrCl3(THF)3 as chromium source, triethylaluminum as activator, and (Ph)2P—N(i-Pr)—P(Ph)-N(i-Pr)—H as ligand for the catalytically active complex as shown below
where Ph is a phenyl group and i-Pr is an isopropyl group This ligand features the typical PNPN—H-backbone, which is why this class of compounds, regardless of the precise nature of its substituents, is often referred to as a “PNPNH-ligand”.
WO 2009/006979 A2 describes essentially modified catalyst systems of the general type already disclosed in EP 2 239 056 B1. These modified systems take advantage from the same PNPNH-type ligands. However, now a “modifier” is added to the system, (but not limited to) ammonium or phosphonium salts of the type [H4E]X, [H3ER]X, [H2ER2]X, [HER3]X or [ER4]X (with E=N or P, X=Cl, Br or I and R=alkyl, cycloalkyl, acyl, aryl, alkenyl, alkynyl etc.).
Preferred embodiments of the invention disclosed in WO 2009/006979 A2 involve, for instance, modifiers such as tetraphenylphosphonium chloride, tetraethylammonium chloride—monohydrate, triethylamine—hydrochloride etc. Also, as a “type [ER4]X”—modifier, dodecyltrimethylammonium chloride can advantageously be used, due to its low price, abundant supply and good solubility in the reaction solution.
In fact, the specifically designed coordination behaviour of the PNPNH ligands is largely the origin of the high selectivities of the catalytically active chromium complexes. Clearly, the high product selectivities are of great importance for the economic viability of the technical process.
Of course, a high selectivity directly results in a minimization of undesired side products in the technical oligomerization process. It is therefore evident that the “key ingredients” of the catalyst have to be produced on technical scale with the highest possible quality.
The laboratory procedure for the preparation of the PNPNH ligand, as demonstrated in Example 1 below, gives a material of good quality. Hexane is assumed to be a good recrystallization solvent due to its non-polar properties.
Using the ligand from the laboratory bench-scale synthesis in standardized catalytic tests of the ethylene trimerization to 1-hexene, it is easily possible to obtain overall 1-hexene yields of 91-93 weight percent at 1-hexene purities of 99.0-99.3% with hardly any detectable wax/polymer formation.
While being transferred to technical scale, however, this laboratory procedure regularly needs some modifications so as to meet the requirements imposed by boundary conditions in a technical environment. For example, in order to avoid hot spots in the reaction mass, it might be advisable to change the dosing sequence and/or dosing speed of some of the ingredients. Furthermore, reaction temperatures as low as −40° C. will, most likely, turn out to be unfavourable or even not feasible on technical scale. Moreover, solvents may have to be recycled.
Even after optimization of the ligand's production process on technical scale, it does not seem to be possible to reach a ligand quality, i.e., purity, comparable to the product synthesized using the laboratory procedure.
One of the most severe problems in all known technical-scale oligomerization processes is the formation of long-chain by-products such as waxes and polyethylene. Clearly, this leads to frequent fouling of equipment such as reactor inner surfaces, heat exchangers, etc. Moreover, wax or polymer formation can lead to plugging of tubing, valves, pumps, and other equipment, making frequent plant shut downs for purging/cleaning and maintenance of equipment necessary.
The measured formation rate of waxes/polymers has to be considered in the design of a commercial ethylene oligomerization plant. Adequate minimization measures and handling procedures for these undesired by-products are inevitable in order to allow for commercially successful plant operation.
Having in mind that, as already pointed out above, a high selectivity results directly in a minimization of undesired side products in this technical process, the “key ingredients,” i.e. especially the ligand, has to be produced on technical scale with the highest possible quality.
The attempt to purify crude PNPNH compound by vacuum distillation using a thin-film evaporator turned out to be rather unsuccessful, since there was hardly any separation effect between the ligand and the impurities.
There accordingly remains a need in the art for a method for purifying a crude PNPNH compound (ligand). Preferably the method is easy to perform with only few process steps.